New type of CataloguedResource: Data product

[jakubklimek]

Currently, in DCAT-AP, we have Datasets, Dataset Series, and Data Services as distinct types of Catalogued resources. However, there is a need identified to also catalog Data products, such as applications, visualizations, or reports based on Datasets, Dataset Series, and Data Services.

Examples:

• In StatDCAT-AP, data visualizations are modeled as Distributions with a special VISUALIZATION type, a practice already identified as breaking the principle that Distributions of a Dataset should have equal content, only accessible in different representations, like different data formats. There is a request for guidance on how to properly represent visualizations and other data products in connection to the Datasets they visualize, or are based on. It is deemed as not specific to statistics, and therefore lifted here.
• In the Czech National Open Data catalog - applications section, there is a list of applications linked to the datasets they use - promoting the reuse of the datasets. Currently, these are modeled as dcat:Resources with a Czech-specific subclass corresponding to "a product or service using open data". The dataset has dcterms:isReferencedBy property pointing to the product. But surely, this is not a Czech-specific need.

Related issue: #202

[bertvannuffelen]

The term Data Product is used in data spaces as the "shared data with its legal, contractual, ... information included": thus the metadata + the data. This is a CEN/CENELEC JTC 25 definition (under review). There is also an OMG standard which calles DataProducts (https://ekgf.github.io/dprod/#dataproductshape) "_A rational, managed, and governed collection of data, with purpose, value and ownership, meeting consumer needs over a planned life-cycle. A data product may have input and output ports, code and metadata" (work inspired/based on financial data management approach for stock exchange purposes).

Given this context I would not use the term data product at all in DCAT-AP (specifications). With data product we enter the world of "marketed data". Which can have the same name as the core dataset, but also can have a more compelling name. In the core it might be the same, but in the context (legal; contractual; expectations) very different. I believe that we can define this markted

The list of examples are for me "consumers" of data. (a report consumes one or more datasets, a visualisation consumes one of more datasets, an application consumes one or more datasets). Based on this maybe we could capture the collection under the title DataConsumer .